Men and their families facing a prostate cancer diagnosis need information and support as they sort through key decisions and adjustments they must make over the subsequent weeks, months and years. The disease and its treatments have substantial and long-lasting effects on quality of life, and our goal is to reduce those effects and speed the process of regaining control over one's quality of life. CHESS (the Comprehensive Health Enhancement Support System) is a Web-accessed system of resources that has been accepted and used, and has successfully improved quality of life for patients with a variety of serious health conditions. The research proposed here would add depth and needs-assessment-based focus to CHESS, make it a system capable of delivering individually-tailored information and resources, and enhance its usefulness for the partners of patients. In addition, we will create procedures for a human Cancer Information Mentor (based on the CIS Information Specialist role) to have a continuing information- supportive relationship with a patient. This project follows and parallels one for breast cancer patients, allowing both substantial efficiencies in development and operation, and the opportunity to compare the two experiences. 360 prostate cancer patients from three hospitals will be randomly assigned to four conditions in a crossed design: either 1) CHESS or open Internet access, and 2) a personal Cancer Information Mentor or not. Where the patient has a partner (often a spouse) interested in participating, that partner will also be offered a separate codename and password for CHESS/internet access. We expect that both CHESS and a Mentor will produce better quality of life on each of four dimensions than accessing the Internet alone, and that the two together will multiply each other's strengths, producing greater benefits than either alone. Drawing on Self-Determination Theory, these effects are expected to occur due to CHESS and Mentor effects on autonomy, competence, and relatedness, each of which is assessed here with two mediating variables. Additional process analyses will distinguish the selectively-variable nature of CHESS and Mentor content and immediate outcomes such as knowledge of one's condition, information overload, etc.